Author: Gary Yang , Founder of Xinghan Capital
Written in Singapore on June 8, 2026
The singularity has accelerated the evolutionary clock of AI, leading to the rapid formation of new generations of civilization across different regions of the world. In the past two months, I participated in over 20 AI-related events in more than ten cities globally, but the Stripe Sessions in downtown San Francisco at the end of April far surpassed all other topics, highlighting the striking generational gap. While the world is struggling with the limitations of single-machine Claws & Agents, Silicon Valley and San Francisco have already entered the next dimension in the management of Agent economics and Agent epistemology. The competitive pressure in Q3 and Q4 of 2026 remains intense, with a very steep exponential curve.
1. Competition in AI Payments and Bottlenecks in the H2A Economy
In Q1, we predicted that the competition for AI Agent Payments would intensify rapidly in many parts of the world during April and May. The demand for value exchange through agents was beginning to emerge, and the rapid development of AI Payments was confirmed in Q2. Following x402, multiple AI Payment Protocols, such as MPP, emerged rapidly in Q2. Not only were traditional and crypto financial payment companies upgrading to AI at full speed, but large companies (especially Google) and even established IT companies (such as IBM) also rushed into this field, hoping to seize the initiative in the world of agents.
On the day of the Stripe Sessions in San Francisco, I discussed the standardization and application of Payment Protocols with technical leaders from several top AI companies. The results were reasonable but not entirely satisfactory:
- ① No one can set the standards; consensus standards can only be gradually formed during the process of vying for dominance.
- ②Most people fully agree that Crypto is the inevitable AI Payment Protocol, but they all start with the Fiat API, partly due to inertia but more so due to compliance obstacles;
- ③KYC is both unavoidable and contrary to Agent Native;
- ④ Everyone is claiming to be A2A (Agent to Agent), and everyone is doing H2A (Human to Agent).
In fact, in Q2 2026, many large and mid-tier companies in Silicon Valley were very similar to companies in East Asia. Even most department heads at the Mag 7 level were still using the AI Payment and Agent Economy hype with a B2C business objective, setting KPIs for lower-level employees based on human users. This inevitably led to the current temporary unorthodox nature of the Payment Protocol and A2A economy. This H2A-oriented trend quickly reached a bottleneck in Q2. The reason is simple: the biggest characteristic of AI agents is their ability to make decisions; however, in the context of internet development, both B2C commerce and the H2A economy are essentially driven by human decision-making. Using agents to help people make payments in traditional e-commerce scenarios is logically non-AI-native, so at this stage, its hype value outweighs its practicality.
However, from another perspective, H2A did indeed play a very good role in stimulating the transition to the next stage of thinking about AI-Native and Agent Autonomous economies. By the end of Q2 2026, some smart companies had realized this and began to "openly repair the plank road while secretly crossing the Chencang pass" by using the AI-Native Agent economy mindset to think about the problem in reverse, and to deduce that the current H2A economic interface was the best value for Q2-Q3.
2. The Inevitable Trend of Agent Economy and A2A Ecosystem
Agent economy refers to a new economic system in which autonomous AI agents directly participate in value creation, value exchange, and value capitalization, and gradually become independent economic entities.
The A2A ecosystem is a process in which different agents participate in economic activities in the agent economy, face each other, and interact (exchange information) (value) to form an overall picture of competitive and cooperative economic value.
In Q2 2026, several top global venture capital firms declared their emphasis on investing in the Agent Economy and A2A ecosystem, even defining it as the only important investment direction for the next stage.
Similar to the pre-e-development phase of e-commerce in 2007, mobile internet in 2013, and Crypto DeFi in 2019, the construction of Agent Economy and A2A ecosystem also requires technical standards, economic rules, consensus building, and market education. While the paradigms are essentially the same, the differences are: ① This time, the fundamental technology is iterating much faster; ② The perspectives of "to A" and "to B" and "to C" differ, not entirely based on human perspectives and needs, being more abstract and difficult to understand, requiring stronger support from first principles, and needing to consider energy consumption and operational efficiency issues from an AI-native perspective; ③ Due to the conflict between the first two points, coupled with regional biases and compliance issues, short-term consensus is more difficult to achieve. The terrible thing is that the evolution of AI will not slow down due to these problems. In other words, the formation of Agent Economy and A2A ecosystem is essentially gradually moving away from human-defined rules and demand frameworks; for them, it's more about breaking through a few quantifiable bottlenecks.
This is a game of rapidly shifting equilibrium. The explosive growth of AI protocols in Q2 2026 fully illustrates this point. Major companies and frontier labs are vying for the entry-level rules of AI agents, and the initial infrastructure of the agent economy is taking shape, like a draft version of the Code of Hammurabi. The equilibrium of traditional finance and commerce will be rapidly dismantled and reshaped in this paradigm shift. Whoever can quickly understand the protocol-based thinking of AI-native and implement it to gain a differentiated advantage will be able to share the AI pie in this game of shifting equilibrium.
3. The connection, gap, and political and economic factors between AI Protocols and Crypto Protocols
AI Protocol is the infrastructure for AI Agents to participate in the Agent Economy. It is also the basic rules, standards and consensus mechanisms that enable Agents to discover, communicate, exchange, and collaborate in economic activities within an Open Network. Simply put, it is the governance rules and economic law of the AI world.
I started writing the AI Protocol at the end of Q1 2026. Initially, it was like a primitive man with hunting experience suddenly entering modern society and participating in the formulation of business rules. It wasn't until I met a Google executive that my team and I quickly got on the right track. The formation and maturation of the AI Protocol carried the aesthetic inertia of large internet companies, but at the same time, it must also follow the first principles of the future AI ecosystem.
The encapsulation forms of AI protocols are currently quite inconsistent, typically taking the form of files (.json, .ts, .txt), CLI, or API/SDK, which is very different from crypto protocols. Firstly, in the early stages of AI development, many communication trust handshakes have not yet established universal standards. Secondly, the content exchanged between AI protocols and crypto protocols at this stage differs. The former involves information gaps, capability gaps, and computing power gaps that need to be exchanged but whose boundaries are not yet clear, while the latter involves relatively clearer boundaries of asset rights, ownership, and governance rights.
A sharp and obvious question arises: Are AI Protocol and Crypto Protocol the same thing? Will they merge into one in the future? I cannot prove this conjecture mathematically at the moment, but intuitively, they will gradually merge and most of their parts will overlap to form a mature Digital Protocol system.
There's a deeper, hidden issue: AI Protocols, at their current stage, prioritize establishing communication and collaboration while downplaying financial governance power and blurring boundaries. This contrasts sharply with Crypto Protocols' philosophy of establishing systems, defining rights, and defining value. The gap is so significant that it could be considered two entirely different philosophies. Besides the surface factor that the AI Agent economy is in its early stages of development and its entry point differs from Crypto Protocols, what other hidden factors contribute to this phenomenon?
Yes, it's very clear: political and economic factors. The major economies and regions of the world, due to their traditional financial and legal compliance foundations, are strongly influencing this gap. In other words, the current AI Protocol and Agent economy still operates within the previous paradigm of human society. All protocols related to money and management are passively avoided, or temporarily and compensatorily constrained by the governance habits of traditional financial and legal systems (Note 1). However, as the energy of this gap accumulates, compared to the exponentially rapid development of AI, an irreconcilable situation will soon emerge, as I summarized at a Cambridge CJBS conference last month:
"AI agents do not think according to the inertia of human society, nor are they motivated to follow the compliance habits of traditional finance. In the next decade, most global financial laws will become ineffective or face severe challenges because AI agents only follow:"
1. First Principles
2. The principle of shortest path for energy value and the principle of highest efficiency
3. Effective KYA, not KYC that conforms to past aesthetics.
The trend of AI Protocols converging into Crypto Protocols is an inevitable consequence of first principles.
4. Paradigm Analogy between AI Agent Submicroeconomics and Biology
The term "AI Agent sub-microeconomics" was first used by me in a recent discussion with an AI expert friend in Oxford, and it has gradually appeared more frequently in our communications with partners over the past two weeks.
Whether the current trend is called the AI economy or the Agent economy, we find that they differ somewhat from human economics in their behavioral characteristics. While they share some paradigmatic comparability, they are not entirely the same. Below, I will briefly outline some differences between the AI Agent economy and the human socio-economic system:
①AI Agents interact with transactions more frequently, but the amount per transaction is lower;
②The consumption and exchange of economic value by AI Agents are more directly related to energy;
③ AI agents make decisions based on efficiency rather than emotion;
④ The economic behavior of AI agents is task-oriented rather than consumption-oriented;
⑤ The organizational cost and marginal learning cost of AI agents approach zero;
⑥ The value consensus of AI Agents is based on communication protocols, and the communication wear and tear costs are almost zero;
⑦ The smallest economic entity and the smallest value unit in the AI Agent economy are different, and can be compared to biology.
In fact, these are just some of the differences that can be seen or foreseen at present. In the future development of AI, there will certainly be many more differences in the derivatives and processes.
The last of the above distinctions, the analogy to biology, is the cornerstone idea that has been most helpful to our business development since Q2 2026, and it is also the most effective model for thinking about products, markets, and management methods from the perspective of AI company commercialization. Specific analogies are as follows:
①LLM, as the driving kernel of agent thinking, is similar to the cell nucleus;
②Agent Harness brings about differentiation in agent operational capabilities, similar to cytoplasm;
③The Agent as a whole is a governance unit with independent task capabilities, possessing subjectivity and functional specificity, similar to a cell;
④ The information communication boundary of an agent is usually a set of network protocol stacks, similar to the conditional passage of substances allowed by the phospholipid bilayer of a cell membrane;
⑤ Value systems and environments outside of the Agent, such as Skills, Prompts, Algorithms, Cli, and the increasingly emerging Composite Skills, Skill Factories, etc., are similar to the extracellular environment, including exosomes, tissue fluid, extracellular matrix, exchangeable nutrients, and various metabolic environments.
In the development and iteration of Q1-Q2 2026, AI Agents are gradually forming clearer boundaries, clearer subjectivity, and clearer principles for the exchange of information, value, and energy. A sub-microeconomic environment for AI Agents, similar to the environment of a biological organism, is taking shape. This environment contains a wealth of AI and economic value that can be explored, and AI Protocols and AI Finance are inevitable trends poised for explosive growth.
5. The inevitability of AIFi and the economic significance of FinChip (financial chip)
Starting in the second half of last year, we began to think about and plan in the direction of AIFi (Artificial Intelligence in Finance). By the end of Q1 2026, the concept of AIFi had formed a clear trend. If we were to give AIFi a relatively clear definition, it could be: the financial system and infrastructure formed by the exchange, trading, and capitalization of the inherent value of AI in the agent economy.
The biggest difference between AIFi and DeFi and TraditionalFi is that the value of DeFi and TraditionalFi is inherent in Fi (Finance), while Decentralized and Traditional are forms of value; AIFi, on the other hand, has value in AI, but Fi becomes the form of value. This is not a simple word game, but the result of a qualitative leap in the development of AI.
Simply put, AI used to serve quantitative strategies, financial products, and production processes; it was merely a development tool for extracting financial and production value. Now, the decision-making capabilities of AI agents have transferred the ability and power of value discovery from people and companies to the agents. The subject of the economic unit has shifted, so the subject of value has also undergone a fundamental change.
Under this trend, building the infrastructure for a new value system will be a crucial task. In my previous article in February, <AI-Fi Financial Chips and Global Finance After the OpenClaw Singularity> ( related reading: AI-Fi Financial Chips and Global Finance After the OpenClaw Singularity: How to Avoid Being Left Behind? ), I first introduced the concept of financial chips (FinChip) and mentioned that the super-intelligent financial assets encapsulated by AI Agent + Crypto Smart Contract will truly adapt to the development of the next era of AI Agent economy. After three months of iterative upgrades, FinChip.AI has initially developed an independent AI Autonomous + Crypto Protocol AIFi system, compatible with both H2A and A2A dual-phase environments; building the infrastructure for the AI Agent economy in an Open Network and gradually forming AI financial value is the important economic significance of FinChip .
6. AI-Native represents a paradigm shift different from "Internet Plus".
Whether it's AIFi, financial circuit principles (Note 2), or financial chips (FinChip), the most important thing is to natively integrate the essential principles of AI, Crypto, and Finance to form a reasonable value system and management mechanism from a future perspective. AI-Native Thinking is the abstract and counterintuitive logic at this stage. As mentioned earlier, "AI follows first principles, as well as the shortest path principle of energy value and the principle of maximum efficiency." This is the most important core challenge for current thinking and the construction of new business paradigms.
Back in February of this year, when OpenClaw was driving this wave of AI upgrades, I and several entrepreneurs discussed a prediction: the upgrade of enterprises using AI+ will be completely different from the upgrade of enterprises using Internet+.
Due to the rapid development, abstract nature, and deeper integration of AI with various business processes, it will be difficult to develop a set of effective industrial upgrading tools and methodologies or general professional consulting opinions for a considerable period (at least two years). The pressure of a steep curve will always exist, posing a significant challenge to all scientists, engineers, and entrepreneurs, and the process of paradigm shifting will be completely different from any previous experience.
